This project will use novel quantitative imaging methods to guide biopsies to biologically distinct regions of primary and post-treatment recurrent GBM for targeted exome, epigenome and transcriptome analysis. Our goal is to identify naturally evolving and treatment-induced mutations and epimutations that promote the selective outgrowth of malignant subclones over lime. Genomic analysis of cancer is typically conducted at a single time point and on a single piece of the bulk resection without knowledge of its original context within the heterogeneous tumor. In contrast to these traditional genomic studies, an image guided approach to newly diagnosed and recurrent tumors could enrich for the detection of drivers of tumor growth by linking mutations and epimutations to regions of aggressive tumor growth in vivo. We will use innovative metabolic and physiologic imaging to identify regions with different levels of proliferation and hypoxia within the same patient. To our knowledge, this would be the first time that advanced imaging will be used to guide genomic or epigenomic analysis of any human tumor. In Aim 1, we will identify functional mutations and epimutations that exhibit intratumoral heterogeneity within newly diagnosed GBM. In Aim 2, we will identify functional mutations and epimutations commonly acquired during tumor progression using image guided tissue samples from treated, recurrent GBM, including paired samples from individual patients over time. Our preliminary data show that chemotherapy can have a profound effect on selective outgrowth of malignant subclones. The integration of data from Aims 1 and 2 will identify subclones in newly diagnosed tumor that exhibit selective outgrowth to become the dominant clone(s) at recurrence, and the sequential biallelic events involving intersecting genetic and epigenetic mechanisms that contribute to their enhanced growth potential. Candidate driver alterations will be evaluated using a mature computational pipeline, and will experimentally be tested for predicted functional effect. These studies could therefore impact patient care by the identification of common drivers specific to recurrence, defining the influence of therapy on tumor evolution, and incorporating profiles of primary and recurrent tumors into personalized treatment plans.